Adaptive QFT control using hybrid global optimization and constraint propagation techniques

Abstract

We propose a procedure for the online design of adaptive quantitative feedback theory (QFT) control system. The proposed procedure uses hybrid global optimization and interval constraint propagation techniques to automatically design online an adaptive QFT controller and prefilter as and when required. While the hybrid global optimization combines interval global optimization and nonlinear local optimization methods, the interval constraint propagation techniques accelerate the optimization search by very effectively discarding infeasible controller parameter regions. The proposed adaptive QFT control is experimentally demonstrated on a coupled tanks system in the laboratory. Experimental results show the superiority of the proposed adaptive QFT control over standard QFT control, in terms of both reduced error and reduced control effort.